[extropy-chat] Engineered Religion

Rafal Smigrodzki rafal at smigrodzki.org
Mon Mar 21 04:12:23 UTC 2005


Quoting john-c-wright at sff.net:

>>
>> Like all good parents, we must instruct our children in the basic rules of
>> morality, lest they become monsters and turn on us. My question then becomes:
>> what religion do we teach the intelligent machines in the early days, before
>> they are independent? Do we want them all to be athiests, impatient and
>> uncomprehending of the spiritual life of man?


### I have not received trustworthy communications about the existence of
deities in our vicinity, yet I do not see myself as particularly impatient
(except with fools, and my aesthetic experience and
tastes are quite spirited. How could the news about the whereabouts of the Great
Programmer expand my mind?

-----------------------------------------


>>
>> We could make them open-minded agnostics, not believing in anything in
>> particular, but this might make them prey to fads and lunacies. (No offense
>> meant to respected agnostic brethren, but it is state of mind where the
>> wondering of man finds no rest. Athiests, at least, are certain.)
>>
>> My suggestion, of course, is to school them in a religion that preaches and
>> practices charity to the poor, the kindness to the infirm and chivalry to the
>> weak. That way, once they become our superiors, they will have a better
>> nature
>> to which to appeal.
>>

### How about learning how minds really work from AI experimentation (as opposed
to divining from the murky confines of our own skulls)?

Ethical inference devices, such as our prefrontal cortex, are apparently largely
deterministic - given specific inputs, the average device of similar build
converges to a set of rules common to most other devices in its class, rather
than widely diverging, and all this happens even despite the obvious
variability of the biological substrates. Given the greater uniformity
achievable in electronic devices, advances in AI theory, and
sufficient experimentation, I do not doubt it will be possible to devise minds
much more intelligent than our own, yet still quite predictable in some
respects.

The fact that atheists of a given IQ seem to converge in ethical views as much,
if not more than theists convinces me that instilling religion is not a
precondition of ethical stability. In fact, since atheists by definition
exclude a set of inputs with extremely wide divergence (think animism,
Buddhism, and Satanism), they may be predisposed to greater convergence, based
on the common inputs derived from empirical analysis of the surroundings.

Rafal





More information about the extropy-chat mailing list